Add tests for encrypted disk and codec and ReplicatedMergeTree. (#30172)

This commit is contained in:
Vitaly Baranov 2021-10-22 14:09:47 +03:00 committed by GitHub
parent 151daf61c1
commit c0d295980d
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
16 changed files with 329 additions and 0 deletions

View File

@ -0,0 +1,9 @@
<clickhouse>
<storage_configuration>
<disks>
<disk_encrypted>
<key>aaaaaaaaaaaaaaaa</key>
</disk_encrypted>
</disks>
</storage_configuration>
</clickhouse>

View File

@ -0,0 +1,9 @@
<clickhouse>
<storage_configuration>
<disks>
<disk_encrypted>
<key>bbbbbbbbbbbbbbbb</key>
</disk_encrypted>
</disks>
</storage_configuration>
</clickhouse>

View File

@ -0,0 +1,16 @@
<clickhouse>
<remote_servers>
<cluster>
<shard>
<replica>
<host>node1</host>
<port>9000</port>
</replica>
<replica>
<host>node2</host>
<port>9000</port>
</replica>
</shard>
</cluster>
</remote_servers>
</clickhouse>

View File

@ -0,0 +1,25 @@
<clickhouse>
<storage_configuration>
<disks>
<disk_local>
<type>local</type>
<path>/disk/</path>
</disk_local>
<disk_encrypted>
<type>encrypted</type>
<disk>disk_local</disk>
<path>encrypted/</path>
<key>0000000000000000</key>
</disk_encrypted>
</disks>
<policies>
<encrypted_policy>
<volumes>
<main>
<disk>disk_encrypted</disk>
</main>
</volumes>
</encrypted_policy>
</policies>
</storage_configuration>
</clickhouse>

View File

@ -0,0 +1,87 @@
import pytest
from helpers.cluster import ClickHouseCluster
from helpers.test_tools import assert_eq_with_retry, TSV
import os
SCRIPT_DIR = os.path.dirname(os.path.realpath(__file__))
cluster = ClickHouseCluster(__file__)
node1 = cluster.add_instance("node1",
main_configs=["configs/remote_servers.xml", "configs/storage.xml"],
tmpfs=["/disk:size=100M"],
macros={'replica': 'node1'},
with_zookeeper=True)
node2 = cluster.add_instance("node2",
main_configs=["configs/remote_servers.xml", "configs/storage.xml"],
tmpfs=["/disk:size=100M"],
macros={'replica': 'node2'},
with_zookeeper=True)
@pytest.fixture(scope="module", autouse=True)
def start_cluster():
try:
cluster.start()
yield
finally:
cluster.shutdown()
def copy_keys(instance, keys_file_name):
instance.copy_file_to_container(os.path.join(SCRIPT_DIR, f"configs/{keys_file_name}.xml"), "/etc/clickhouse-server/config.d/z_keys.xml")
instance.query("SYSTEM RELOAD CONFIG")
def create_table():
node1.query("DROP TABLE IF EXISTS tbl ON CLUSTER 'cluster' NO DELAY")
node1.query(
"""
CREATE TABLE tbl ON CLUSTER 'cluster' (
id Int64,
str String
) ENGINE=ReplicatedMergeTree('/clickhouse/tables/tbl/', '{replica}')
ORDER BY id
SETTINGS storage_policy='encrypted_policy'
"""
)
def insert_data():
node1.query("INSERT INTO tbl VALUES (1, 'str1')")
node2.query("INSERT INTO tbl VALUES (1, 'str1')") # Test deduplication
node2.query("INSERT INTO tbl VALUES (2, 'str2')")
def optimize_table():
node1.query("OPTIMIZE TABLE tbl ON CLUSTER 'cluster' FINAL")
def check_table():
expected=[[1, 'str1'], [2, 'str2']]
assert node1.query("SELECT * FROM tbl ORDER BY id") == TSV(expected)
assert node2.query("SELECT * FROM tbl ORDER BY id") == TSV(expected)
assert node1.query("CHECK TABLE tbl") == "1\n"
assert node2.query("CHECK TABLE tbl") == "1\n"
# Actual tests:
def test_same_keys():
copy_keys(node1, 'key_a')
copy_keys(node2, 'key_a')
create_table()
insert_data()
check_table()
optimize_table()
check_table()
def test_different_keys():
copy_keys(node1, 'key_a')
copy_keys(node2, 'key_b')
create_table()
insert_data()
check_table()
optimize_table()
check_table()

View File

@ -0,0 +1,7 @@
<clickhouse>
<encryption_codecs>
<aes_128_gcm_siv>
<key>0000000000000000</key>
</aes_128_gcm_siv>
</encryption_codecs>
</clickhouse>

View File

@ -0,0 +1,7 @@
<clickhouse>
<encryption_codecs>
<aes_128_gcm_siv>
<key>aaaaaaaaaaaaaaaa</key>
</aes_128_gcm_siv>
</encryption_codecs>
</clickhouse>

View File

@ -0,0 +1,10 @@
<clickhouse>
<encryption_codecs>
<aes_128_gcm_siv>
<key remove="1"/>
<key id="0">aaaaaaaaaaaaaaaa</key>
<key id="1">bbbbbbbbbbbbbbbb</key>
<current_key_id>0</current_key_id>
</aes_128_gcm_siv>
</encryption_codecs>
</clickhouse>

View File

@ -0,0 +1,10 @@
<clickhouse>
<encryption_codecs>
<aes_128_gcm_siv>
<key remove="1"/>
<key id="0">aaaaaaaaaaaaaaaa</key>
<key id="1">bbbbbbbbbbbbbbbb</key>
<current_key_id>1</current_key_id>
</aes_128_gcm_siv>
</encryption_codecs>
</clickhouse>

View File

@ -0,0 +1,8 @@
<clickhouse>
<encryption_codecs>
<aes_128_gcm_siv>
<key>aaaaaaaaaaaaaaaa</key>
<nonce>xxxxxxxxxxxx</nonce>
</aes_128_gcm_siv>
</encryption_codecs>
</clickhouse>

View File

@ -0,0 +1,8 @@
<clickhouse>
<encryption_codecs>
<aes_128_gcm_siv>
<key>aaaaaaaaaaaaaaaa</key>
<nonce>yyyyyyyyyyyy</nonce>
</aes_128_gcm_siv>
</encryption_codecs>
</clickhouse>

View File

@ -0,0 +1,7 @@
<clickhouse>
<encryption_codecs>
<aes_128_gcm_siv>
<key>bbbbbbbbbbbbbbbb</key>
</aes_128_gcm_siv>
</encryption_codecs>
</clickhouse>

View File

@ -0,0 +1,16 @@
<clickhouse>
<remote_servers>
<cluster>
<shard>
<replica>
<host>node1</host>
<port>9000</port>
</replica>
<replica>
<host>node2</host>
<port>9000</port>
</replica>
</shard>
</cluster>
</remote_servers>
</clickhouse>

View File

@ -0,0 +1,110 @@
import pytest
from helpers.cluster import ClickHouseCluster
from helpers.test_tools import assert_eq_with_retry, TSV
import os
SCRIPT_DIR = os.path.dirname(os.path.realpath(__file__))
cluster = ClickHouseCluster(__file__)
node1 = cluster.add_instance("node1",
main_configs=["configs/remote_servers.xml", "configs/encryption_codec.xml"],
macros={'replica': 'node1'},
with_zookeeper=True)
node2 = cluster.add_instance("node2",
main_configs=["configs/remote_servers.xml", "configs/encryption_codec.xml"],
macros={'replica': 'node2'},
with_zookeeper=True)
@pytest.fixture(scope="module", autouse=True)
def start_cluster():
try:
cluster.start()
yield
finally:
cluster.shutdown()
def copy_keys(instance, keys_file_name):
instance.copy_file_to_container(os.path.join(SCRIPT_DIR, f"configs/{keys_file_name}.xml"), "/etc/clickhouse-server/config.d/z_keys.xml")
instance.query("SYSTEM RELOAD CONFIG")
def create_table():
node1.query("DROP TABLE IF EXISTS tbl ON CLUSTER 'cluster' NO DELAY")
node1.query(
"""
CREATE TABLE tbl ON CLUSTER 'cluster' (
id Int64,
str String Codec(AES_128_GCM_SIV)
) ENGINE=ReplicatedMergeTree('/clickhouse/tables/tbl/', '{replica}')
ORDER BY id
"""
)
def insert_data():
node1.query("INSERT INTO tbl VALUES (1, 'str1')")
node2.query("INSERT INTO tbl VALUES (1, 'str1')") # Test deduplication
node2.query("INSERT INTO tbl VALUES (2, 'str2')")
def optimize_table():
node1.query("OPTIMIZE TABLE tbl ON CLUSTER 'cluster' FINAL")
def check_table():
expected=[[1, 'str1'], [2, 'str2']]
assert node1.query("SELECT * FROM tbl ORDER BY id") == TSV(expected)
assert node2.query("SELECT * FROM tbl ORDER BY id") == TSV(expected)
assert node1.query("CHECK TABLE tbl") == "1\n"
assert node2.query("CHECK TABLE tbl") == "1\n"
# Actual tests:
def test_same_keys():
copy_keys(node1, 'key_a')
copy_keys(node2, 'key_a')
create_table()
insert_data()
check_table()
optimize_table()
check_table()
def test_different_keys():
copy_keys(node1, 'key_a')
copy_keys(node2, 'key_b')
create_table()
insert_data()
assert "BAD_DECRYPT" in node1.query_and_get_error("SELECT * FROM tbl")
assert "BAD_DECRYPT" in node2.query_and_get_error("SELECT * FROM tbl")
# Hang?
#optimize_table()
#check_table()
def test_different_current_key_ids():
copy_keys(node1, 'key_a_and_b_current_a')
copy_keys(node2, 'key_a_and_b_current_b')
create_table()
insert_data()
check_table()
optimize_table()
check_table()
def test_different_nonces():
copy_keys(node1, 'key_a_and_nonce_x')
copy_keys(node2, 'key_a_and_nonce_y')
create_table()
insert_data()
check_table()
optimize_table()
check_table()